International Journal of Wireless and Ad Hoc Communication

Industrial wireless sensor networks are central to the continuous monitoring of critical plant equipment, yet reliable identification of multiple concurrent fault modes from heterogeneous multivariate sensor streams remains an unsolved operational challenge. Physical failure mechanisms—pump cavitation, valve blockage, gradual sensor drift—and wireless channel disturbances each imprint distinct but overlapping temporal signatures that render classical thresholdand rule-based detectors inadequate for automated maintenance dispatch. This paper  presents TA-FaultNet, a neural architecture designed specifically for the multi-class fault identification problem in industrial sensor deployments. The network couples a two-stage stacked bidirectional recurrent encoder with a parallel multi-head self-attention module and a compact temporal convolutional block, enabling simultaneous capture of long-range process dynamics and fine-grained fault-onset localisation from raw sensor windows. TA-FaultNet is evaluated on the publicly available Skoltech Anomaly Benchmark under five operational classes and assessed through a comprehensive battery of experiments including baseline comparisons, systematic component ablation, cross-experiment generalisation, andprogressive noise-injection testing. The proposed architecture decisively outperforms eight competing methods spanning classical anomaly detectors, standalone recurrent and convolutional networks, and the Transformer, while remaining lightweight enough for edge gateway deployment. Attention weight visualisations expose fault-specific temporal activation patterns, providing maintenance engineers with interpretable diagnostic evidence beyond bare classification labels.

The application of quantum key distribution, satellite communication and programmable wireless access in future secure wireless networks is anticipated to enhance secure communication infrastructure. But their realisation demands more than just the physical realisation of quantum links. The network controller needs to determine when a quantum-secured wireless link can be used to serve a request, how orbit type and weather conditions impact the volume of usable keys, and whether the secure key rate is high enough to admit a route. In this paper, we introduce Q-SARA, a quantum-secure access and routing admission model for satellite-assisted quantum wireless networks. It assesses candidate QKD access links with secure key rate (SKR), quantum bit error rate (QBER), link loss, contact duration, visibility probability and propagation delay. A smaller, pre-processed dataset is derived from the public Satellite-to-Ground QKD SKR dataset, which contains the calculated key performance indicators for Low Earth Orbit, Medium Earth Orbit and Geostationary satellite-to-ground QKD links using the prepare-and-measure and entanglement-based protocols. The empirical analysis examines 7,200 link-level data and assesses Q-SARA across orbit, protocol, optical ground station, elevation, atmospheric, and service classes. The findings demonstrate that link selection based only on key volume can be misleading when assessing service quality, while the multi-criteria score provides better balance between security, visibility and latency. LEO links have better instantaneous key rates, GEO links have better visibility, and MEO links lie in between and can be exploited when link quality and service are taken into account together. The results suggest that quantum wireless access should be considered as a service admission problem rather than a physical-layer key generation problem.

As wireless sensor networks (WSNs) and mobile ad hoc networks (MANETs) are (DoS) attacks has become a critical security concern in mission-critical wireless (DoS) attacks has become a critical security issue. This paper proposes ADML-IDS, an Adaptive Machine Learning Intrusion Detection System that integrates ensemble of Random Forest, XGBoost and Gradient Boosting classifiers using a Flooding, and Scheduling—as well as normal traffic. Flooding, Scheduling and normal traffic. Experiments are conducted on the open-source WSN-DS dataset, which contains 166,000 network observations using the LEACH hierarchical routing protocol with 23 features obtained from NS-2 simulation. The data preprocessing steps include Min- Max normalisation and Synthetic Minority Over-Sampling Technique (SMOTE) to balance classes, and importance-based feature selection to retain 19 features. A rigorous ten-fold crossvalidation strategy is followed. ADML-IDS achieves an overall accuracy of 99.57%, weighted F1-score of 0.9956 and AUC-ROC of 0.9985. AUC-ROC of 0.9985, outperforming each of the sub-classifiers and five state-of-the-art methods. Scalability experiments demonstrate that the accuracy of detection remains above network size reaches 200 nodes, and with a reasonable computational cost. A formal presentation of the energy-aware network model and ensemble decision rule is tables are also included along with a full description of the algorithm tables.

Flying ad hoc networks (FANETs) enable dynamic multi-hop communication in un-manned aerial nodes, but their routing plane is vulnerable to selective forwarding attacks that decrease packet delivery rates while avoiding the sudden effects of denial. This paper proposes a trust-aware routing and detection approach for early detection of grey-holes in ad hoc flying networks. The paper employs an analysis-ready data set based on the public FAN-GHETS24 data set, a new data set for early time-series classification of attacks in FANETs. The Trust-Aware Routing Grey-Hole Detection (TAR-GHD) model uses a com-bination of link quality evidence, route stability, packet consistency and trust dynamics in a lightweight detection layer that can be executed alongside traditional ad hoc routing. A mathematical formulation is given for evidence aggregation, temporal trust evolution, risk assessment and route warning. The empirical study measures the detection of normal, mild, moderate and heavy grey-hole attacks in various node-density, mobility, observation window, and classification settings. The findings demonstrate that trust and packet-loss dynamics offer reliable early indicators of grey-hole attacks, while mobility and route changes make it harder to distinguish normal loss from malicious loss. The best-performed configuration resulted in an F1-score over 0.93 (held-out evaluation), with the most influential features related to packet delivery, forwarding ratio, trust score and drop-rate dynamics. The results highlight lightweight and explainable trust evidence as a viable technique for enhancing the security of wireless ad hoc routing in UAV-assisted applications.

With the rapid expansion of the Internet of Things (IoT), reliable and efficient data transmission has become a critical requirement for large-scale heterogeneous deployments. This paper presents a comprehensive simulation-based performance analysis of three widely adopted IoT transport protocols—UDP, CoAP, and MQTT—under Rayleigh fading channel conditions using a MATLAB-based framework. The study evaluates the transmission of 100 and 1000 data packets under three distinct latency regimes: low, medium, and high. Key performance metrics include end-to-end delay, jitter, packet loss ratio, and throughput. A novel Adaptive Exponential Moving Average (EMA) jitter buffer algorithm is proposed, achieving 57–65% jitter reduction across all tested scenarios. Protocol comparison reveals that UDP achieves the lowest average delay (20 ms under low conditions), while MQTT incurs the highest overhead (+20 ms) due to broker relay. Monte Carlo statistical analysis with 500 simulation runs confirms result convergence within 0.5 ms between 100-packet and 1000-packet scales. The findings provide practical design guidelines for IoT protocol selection and establish a reproducible benchmark for evaluating transport-layer behavior in constrained wireless ad hoc networks.